4
$\begingroup$

Let $C_0(\mathbb{R})$ denote the analytic functions $f : \mathbb{R} \rightarrow \mathbb{R}$.

I wonder whether there a functions $f \in C_0(\mathbb{R})$ with $f \neq 0$, such that there is a constant $C$, with $$\left| \frac{d^kf}{d^kx} \right | \leq C$$ for all $k \geq 0$, and $\frac{d^kf}{dx}$ vanish both at $- \infty$ and $\infty$ for all $k \geq 0$.

$\endgroup$
6
  • 5
    $\begingroup$ What about $(\sin x)/x$? $\endgroup$ Oct 4, 2020 at 17:29
  • 4
    $\begingroup$ also, technically, since you didn't specify non-trivial, I guess $f\equiv 0$ is another exampl. $\endgroup$ Oct 4, 2020 at 17:37
  • $\begingroup$ @Willie Wong Thanks for pointing it out, I amended it. $\endgroup$
    – tobias
    Oct 4, 2020 at 17:54
  • $\begingroup$ @RichardStanley It seems $sin(x)/x$ is doing the job. However, is there an easy proof for it? $\endgroup$
    – tobias
    Oct 4, 2020 at 21:10
  • 1
    $\begingroup$ @tobias: for a rescaled sinc function you can also get the proof using my answer. the fact that its Fourier transform is the rectangle function and not smooth is inconsequential: the Fourier transform of all its derivatives are bounded with compact support, so the decay at infinity can be also gotten via Riemann-Lebesgue. $\endgroup$ Oct 5, 2020 at 1:55

1 Answer 1

16
$\begingroup$

Yes.

Let $\phi$ be any smooth function with compact support on the interval $[-1,1]$.

Set $f$ to be the inverse Fourier transform of $\phi$.

Since $\phi$ is in Schwartz class, so is $f$, and all of its derivatives decay to zero as one approach $\pm\infty$.

You can estimate

$$ |f^{(k)}(x) | \lesssim \| |\xi|^k \phi(\xi) \|_{L^1} \leq 2 \|\phi\|_{L^\infty} =: C$$

$f$ is analytic by Paley-Wiener.

$\endgroup$
2
  • 3
    $\begingroup$ Why do you specify “inverse” Fourier transform and not simply the Fourier transform? For thepurposes of question it amounts to the same thing. $\endgroup$
    – KConrad
    Oct 4, 2020 at 18:11
  • 11
    $\begingroup$ Because for my own sanity I like to keep track of which functions are on the frequency domain and which are on the time domain. (Force of habit.) $\endgroup$ Oct 4, 2020 at 18:45

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.